• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 99 subscribers
  • Views 209 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM5728: Skew adjustment in length of RGMII signal line

K.Z
Expert 2380 points
Part Number: AM5728

Tool/software:

Hi All,

I have a question about the signal line length of RGMII of AM5728.
Currently, the line length difference between RXD(0:3) and TXD(0:3) is ± a few millimeters relative to the respective clock signals RXC and TXC.
So, is there any hardware (line length) that should be matched for skew adjustment?
Of course, skew adjustment is also done by software.

Example 1 : Should the line length be longer in the ＋ side direction by adding to CLK ? 
Example 2 : Is it better to keep the difference within 1mm in the 4-bit data of RXD (0:3)?

Best Regards,

ito

  • +1
    TI__Guru 61520 points

    Ito,

    The important thing to keep in mind is that either H.W or S/w methods are fine as long as the timing violations do not occur. It would be ideal to run the simulations of the board when you design them to ensure that the timing budget is not violated.

    The TXC is always internally delayed before sending out the pins. the RXC must be externally delayed. This might help in your case. If the TXC of the PHY is internally delayed, then I think it should work fine.

    Please refer to the RGMII interface timing budgets app note.

    https://www.ti.com/lit/an/snla243/snla243.pdf

  • +1 in reply to ShreyasRao
    TI__Guru 61520 points

    The datasheet mentions that the board prop delays are matched within 50ps of TXC w.r.t data and ctrl.

    From your table, we believe that you have plenty of margin.